The link above has some errors in the MD5 checksum data presented on the GIAB github table. Minor, but thought worth drawing to your attention, if you wanted to verify (and correct?)
On the GIAB github page:
the bam & bai file for HG002 currently both show the same MD5 hash (= c80f0cab24bfaa504393457b8f7191fa).
In my download, that hash matches the .bam file, but the .bai file comes up as d4fea426c3e2e9a71bb92e6526b4df6f
the bai file for HG004 shows MD5 hash = 8914bfb6fa6bd304192f2c9e13e903f.
In my download, the hash comes up as 8914bfb6fa6bd304192f2c9e13e903f4. Close enough to guess that the website text was probably inadvertently truncated by a digit, rather than an actual mismatch.
I just downloaded the Illumina WES data from the Ashkenazi trio: https://github.com/genome-in-a-bottle/giab_data_indexes/blob/master/AshkenazimTrio/alignment.index.AJtrio_OsloUniversityHospital_IlluminaExome_bwamem_GRCh37_11252015
The link above has some errors in the MD5 checksum data presented on the GIAB github table. Minor, but thought worth drawing to your attention, if you wanted to verify (and correct?)
On the GIAB github page: